by Peter Yule
called ‘Dannii’. Although Oscar Hughes as project director always
resisted changes as they would threaten the budget and the sched-
ule, he realised that the new air purifier would benefit the project
and approved the change, even though there was an argument
that it should have been a designer’s change rather than a cost to
the project. Greg Stuart says that: ‘We could have left the scrubber
until they did trials and told Kockums that their system did not
meet endurance specifications, but the new system would have
been hard or impossible to back-fit and . . . the boat was getting
too heavy and the scrubber change helped reduce this.’
Another change made late in the design process was the speci-
fication for electrical cabling. A particular ‘low fire hazard’ cable
had been specified in the contract but later tests showed that it
was a serious fire risk. Greg Stuart recalls going to see Oscar
Hughes and saying that the cable would cause a disaster one day.
Although the project was in a poor position to enforce a change
and it would cause delays and increased weight, Hughes agreed
that it was needed.9
The way the change was implemented showed a degree of
commonsense and flexibility that was perhaps lacking later in the
project. Oscar Hughes recalls:
There was no choice but to change the cables to a less toxic
sheathing material. The challenge lay in the implications in
terms of schedule (a potential delay of 14 months) and cost
D E S I G N I N G T H E C O L L I N S C L A S S
141
on the program as the first submarine was well advanced. As
I recall, the solution reached also involved consideration of
the redesign of the bilge pump . . . ASC agreed to accept
responsibility for the design and procurement of a new bilge
pump, the Commonwealth agreed to fund the cost of the new
cables including small premiums for a shortened
manufacturing lead time and their installation. ASC agreed to
fund the electrical cable redesign work required by Kockums
and other sub-contractors and to maintain the original
delivery schedule as contracted, and the Commonwealth
agreed to waive liquidated damages for late delivery of the
first submarine for up to three months. (The cost to the
project was about) $6m . . . a figure significantly less than a
delay to the program of 14 months noting the safety benefits
of the new cable material and the greatly enhanced
performance of the new bilge pump. A great outcome!10
A significant change that had major consequences for the sub-
marine design came about through the addition of anechoic tiles.
There had always been a requirement that the submarine be able
to be fitted with tiles, but their specifications were unknown until
well into the design process. It turned out that the tiles were sig-
nificantly heavier than expected, so the submarines had to grow
by several hundred tonnes to provide sufficient buoyancy to carry
them. Inevitably this had ramifications for many other aspects of
the design.11
The active design process continued in Malm ö throughout the
late 1980s and early 1990s, even as the site in Adelaide was pre-
pared and construction of the first submarines began. As building
progressed the responsibility for finalising the design was gradu-
ally transferred to Adelaide, where an increasing amount of the
work was done by Australians. As Martin Edwards says, a con-
siderable number of Australians had by then been working on the
design for four or five years and had significant experience – in
some areas more experience with the Collins class than the Swedes
who had originally been teaching them.
C H A P T E R 13
Building submarines
By late 1989 the Australian Submarine Corporation’s new ship-
yard was completed. ASC moved from its temporary premises in
Woodville and construction of the first hull sections began in the
150 metre long workshop.
By that time work was already underway at hundreds of facto-
ries around the world on parts and equipment for the submarines.
At Champagne-sur-Seine on the outskirts of Paris the first propul-
sion motor was being built, and the Westinghouse factory in
Sydney was preparing to build five more. At Hedemora in
Sweden the prototype diesel engine had been built and tested,
and Australian Defence Industries at Garden Island in Sydney was
gearing up to build the remaining engines. Strachan & Henshaw in
Bristol was working closely with Kockums and the manufacturers,
Australian Defence Industries in Bendigo, Victoria, on the design
and construction of the torpedo tubes and weapons discharge sys-
tems. At J önk öping in southern Sweden Saab Instruments was
working with its Australian partner Wormald on the ship con-
trol system – regarded as one of the riskiest areas of the sub-
marine project. At Kockums’ shipyard in M älmo welding crews
were at work on two major sections of the first submarine. At its
142
B U I L D I N G S U B M A R I N E S
143
plant in Blacktown, New South Wales, Chicago Bridge & Iron
had begun fabrication of the steel for sections to be assembled in
Adelaide, while the engineering firm Johns Perry, a subsidiary of
Boral, was beginning the complex manufacture of the platforms
on which most of the equipment would be assembled before being
inserted into the hull sections. Across the road from ASC, the
battery factory of Pacific Marine Batteries was the first of what
was hoped would be a whole industrial suburb of defence-related
factories.
Hundreds of companies in Australia and overseas were begin-
ning work on thousands of smaller items. In Norway Phontech
was making broadcast speakers and magnetic loop amplifiers, and
in Melbourne Ryco Hydraulics was making hoses and compo-
nents. In Denmark Novenco Anderberg was making high pressure
air dryers, and in Brisbane Liset Engineering made steel moulds
for the submarines’ casings. British company Michell Bearings
won the contract for propeller shaft thrust bearings, which were
to be made by Perry Engineering in Adelaide. And with the sub-
contracts came staff from Kockums, ASC and the project office to
teach, advise and ensure that quality standards were met.
Overseas companies began working closely with Australian
companies in training staff and preparing factories to manufacture
components in Australia. For example, British company Marconi
Underwater Systems had a contract for designing acoustic win-
dows and transferring technology to Australia. The first set was to
be made in Britain so Australians could learn the process, with the
later sets to be made in Toowoomba, Queensland, by Buchanan
Aircraft Corporation.
Many of these overseas companies set up Australian sub-
sidiaries to manufacture components in Australia. For example,
Rexroth GmbH of Germany had a contract for hydraulic pumps,
which were made by its Australian subsidiary in Adelaide.
Parts made around the world ranged from tiny fasteners and
clamps to the two complete sections of the first submarine made
by Kockums. These all had to arrive in Adelaide in a systematic
way to be fitted to each submarine as it went through the assembly
process, when the modules from around Australia and the world
were put together. ASC did not have extensive workshops but did
need to have sophisticated project management and integration
skills.
144
T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y
An additional factor in building the Collins class was that it
was approached as a fast-track design and build program, mean-
ing that building began before the design was completed. This
technique is common with many large projects, lowering cost
by reducing the schedule imposed by waiting for the completed
design. However, it carries obvious risks and can put great pressure
on participants.1 Constant design changes were the biggest chal-
lenge faced by many sub-contractors on the submarine project.
Most of the hundreds of sub-contracts went smoothly but,
somewhat ironically, one of the few that caused difficulties was
that with Kockums for the construction of the 300 and 600 sec-
tions of the first submarine. Roger Sprimont recalls that there
were several reasons why Kockums fought hard to build the first
submarine in Sweden. Partly, of course, the company wanted the
work for its own shipyard, but, more importantly, it was seen as
the best way to prove the various production procedures being
developed by Kockums. Also, Kockums felt it was important to
manage Chicago Bridge & Iron and prevent it from taking com-
plete control of the hull production in Australia with procedures
not suited to Kockums’ design philosophy. As it turned out Aus-
tralia resisted having the whole of the first submarine built in
Malm ö, and Kockums agreed to a compromise whereby it would
build the two most complex sections.
The work successfully proved Kockums’ production proce-
dures and transferred them to Australians from ASC. Yet despite
Kockums’ lengthy experience, the welding in these sections proved
to be seriously flawed. And as with so many parts of the Collins
story, different people have totally different versions of what went
wrong, why it went wrong, and how serious it was.2
Greg Stuart was one of the first to raise the alarm. Stuart made
regular visits to Kockums while the two sections were being built
and, talking one day with Lina da Silva, a Brazilian who was
ASC’s inspector at Kockums, was told, ‘I’ve got a problem. I know
Collins is wrong and no one will listen to me.’ The Australians
had free access to the Kockums yard and the two of them went to
look at the sections, where Stuart could see many obvious defects
in the welding. As he succinctly expressed it: ‘With welding, if it
looks like shit, it is shit.’
Fortuitously, Oscar Hughes arrived in Sweden the following
Monday morning and Stuart took him to Kockums as soon as
B U I L D I N G S U B M A R I N E S
145
he got off the plane. They looked at the two sections without
telling anyone, but the Swedes soon heard that the Australian
admiral was in the submarine wearing a hard hat and overalls.
When Hughes and Stuart emerged there was a line-up of senior
Kockums executives waiting for them. Hughes told them: ‘There’s
a big problem and we need to sort it out.’
Stuart recalls that Kockums held a meeting with its welding
engineers and others who said that the welding problems were not
serious. He disagreed strongly, saying that there were hundreds of
visually identifiable defects and he believed that the safety of the
hull was in question.3
Tomy Hjorth, chairman of Kockums and ASC, recalls that
Kockums suggested keeping the sections in Sweden for several
months to repair the defective welds, but Don Williams and Oscar
Hughes were committed to maintaining schedule and asked that
they be sent to Australia where ASC would make the repairs. The
Australians were doing all they could to keep the project moving
and feared that repairing the welds in Sweden would cause lengthy
delays. On the other hand, Mark Gairey saw it as ‘a game call as
at that stage we were not as certain then of the conservative design
and therefore of Collins’ safety’. Nor were they yet certain that
ASC’s welding at Adelaide would reach the required standard.
Accepting the sections with known welding defects involved a
degree of risk.
How did the problem arise? Greg Stuart argues that the main
reason was that the submarines were built with a new steel
alloy developed in Sweden and improved by BHP and Australian
defence scientists, which required different welding techniques
to those normally used by Kockums’ welders. In addition, the
Australian navy required the use of full penetration welds – where
the Swedes had always used partial penetration welds – and a
new type of welding rod. Stuart’s view is that Kockums was over-
confident in the ability of its welders to adapt to the new tech-
niques and did not rigorously supervise and inspect the work.
Those at Kockums generally agree that the welding was
faulty, but contend that its seriousness was exaggerated and used
by the Commonwealth as a bargaining tool after the relation-
ship between Kockums and the Commonwealth collapsed in the
late 1990s. While conceding that their welders did not prop-
erly follow procedures with the new steel and techniques, they
146
T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y
argue the construction was rushed for reasons outside Kockums’
control. The construction schedule for the modules was tight,
and it became tighter because ASC was late in delivering the
steel hull plates. Eventually some of these were flown to Sweden
on chartered jumbo jets, but Kockums was still forced to work
under severe time pressures and this contributed to the welding
problems.
Australian sub-contractors were generally able to reach the
quality standards required for the project, although they some-
times encountered different problems, possibly through their
inexperience with defence contracts. The experience of Boral, a
diversified industrial conglomerate with subsidiaries involved in
several areas of the project, illustrates some of these.
Johns Perry was one of Australia’s major engineering contrac-
tors and, before its takeover by Boral, had been seen as a potential
submarine consortium member. It won a major sub-contract to
build 40 platforms (the first two being built in Sweden) on which
all the interiors of the submarines were constructed.
Manufacture was a highly sophisticated welding exercise. The
platforms have a honeycomb construction designed to resist shock
and minimise vibration, and thei
r production involved welding
hundreds of small components. It proved to be a far greater chal-
lenge than the company had imagined. Bruce Kean, the managing
director of Boral at the time, recalls that they had the technical
capability and systems to do it but underestimated the complexity
of the overall submarine design process. They were overwhelmed
by the myriad changes to design detail and the consequent need
for extensive rework, compounded by the late supply of key com-
ponents from ASC. The need to keep detailed certified records
of the work, including the changes, was a mammoth adminis-
trative exercise in itself. From the contractor’s point of view it
was a nightmare, but ASC finally accepted that costs for design
changes could be recovered. Johns Perry lost money but developed
its quality assurance techniques and learnt many important skills
that proved invaluable in later contracts.
The platforms produced by Perry Engineering were sent to ASC
Engineering for outfitting. ASC Engineering, on the old works of
T. O’Connor & Sons, was a subsidiary of ASC that fitted out
the platforms and manufactured components for the submarines
but also sought other defence and private sector engineering.
B U I L D I N G S U B M A R I N E S
147
Mark Gobell, who had worked for O’Connor since 1973, was
among the staff who transferred after ASC Engineering’s takeover
in 1991. One of the first things that happened was that he and
several others went to Sweden to see how Kockums outfitted the
platforms on their sections of the first submarine. They returned to
prepare ASC Engineering to fit out the platforms when delivered
by Perry Engineering. All the equipment, mechanical components,
piping, electrical components and even the trim and fittings were
attached and some pre-testing of the systems undertaken before
the platforms were transported to ASC for insertion into the sub-
marine.
As with all the Australian sub-contractors, ASC Engineering
had to take special steps to meet the project’s quality standards.
It gained quality certification and trained staff, particularly in the
areas of welding and electrical work, to standards far more rigor-
ous than for civilian work. All sub-contractor work was physically
inspected by the submarine project team. At the project office John